How to choose a Bitcoin mining pool in 2026: the full guide
Pool choice moves your bottom line more than most miners think: with identical hardware, the gap between a good and a bad pick reaches 3-7% net per year - weeks of free hashing lost or gained. This guide walks through every criterion in order: payout scheme, real fee, pool size, payout threshold, stratum latency and reliability. At the end: a step-by-step algorithm and the mistakes we see most often. Live numbers for each pool are in the POOL BTC ranking.
TL;DR checklist
- Payout scheme: FPPS for steady income, PPLNS for lower fees with variance.
- Compare the effective fee (stated fee + hashprice underpayment), not the sticker.
- The payout threshold should clear in 1-2 weeks at your hashrate, not months.
- Ping to the stratum server under 100 ms: every extra hop costs stale shares.
- Check track record: pool age, payout delays, public block stats.
- Run your numbers through the FPPS vs PPLNS calculator.
1. Payout scheme: FPPS, PPS+, PPLNS, TIDES
FPPS (Full Pay Per Share) pays a fixed amount per share, including a slice of network transaction fees. Income is predictable day to day - the pool charges a higher fee for absorbing variance, typically 2-4%. The right fit if you pay monthly power bills and cannot ride out droughts.
PPLNS (Pay Per Last N Shares) distributes actual block rewards. Fees are lower (0.5-2%) but income swings: a week without blocks is a week without payouts. Over 6-12 months a steady miner usually nets 1-2% more on PPLNS than FPPS. The full comparison with numbers is in FPPS vs PPLNS: which payout model to choose.
TIDES (Ocean) is a PPLNS variant with non-custodial payouts straight from the coinbase transaction: the pool never holds your coins. The price is variance and a hashrate floor for payouts.
2. The real fee vs the advertised fee
The advertised "2%" is not the whole price. Pools differ in the payout base: some pay on full hashprice (block subsidy plus network fees), others on subsidy only. With a busy mempool the difference reaches 1-3% of revenue. The check is simple: take one day of payouts, divide by your hashrate, and compare against market hashprice. Watch current difficulty and the next adjustment forecast on the network difficulty page: after a big retarget every pool's payout base shifts at once, so always compare pools over the same window.
A worked example across three pools: TRUSTPOOL vs Foundry vs AntPool fee comparison.
3. Pool size: hashrate, luck, variance
A large pool (Foundry, AntPool - tens of EH/s) finds blocks daily: payouts are smooth and luck averages out fast. A small pool can wait weeks for a block - on PPLNS that is your income swinging. Rule of thumb: the smaller your hashrate and risk tolerance, the bigger the pool you want on PPLNS - or take FPPS and stop thinking about luck entirely.
The flip side is concentration: a pool above 30% of the network is a decentralization risk for Bitcoin. If that matters to you, look at Ocean or Braiins with miner-built block templates (Stratum V2).
4. Payout threshold and frequency
A 0.005 BTC threshold takes roughly two weeks to clear at 100 TH/s - and months at 20 TH/s, during which your money sits with the pool. Small farms should look for thresholds from 0.0005 BTC or scheduled payouts. Mind the withdrawal fee too: some pools cover it, others pass on the network fee.
5. Stratum latency and server geography
Every ~100 ms of latency to the pool server translates into a fraction of a percent of stale shares. Pick a pool with servers on your continent, not just in the US. The test is one command: ping stratum.pool.example before you sign up. If your miners sit in a colocation facility, ask the operator about network routes - more in our miner hosting guide.
6. Reliability and transparency
The minimum due diligence before pointing hashrate anywhere:
- Pool age and incident history: payout delays surface on forums within a minute of searching.
- Public stats: pool hashrate, found blocks and luck visible publicly, not only in the dashboard.
- Verifiable payouts: pool wallet transactions visible in a block explorer.
- 2FA and a withdrawal address whitelist - non-negotiable.
7. Special cases
One or two ASICs at home. Solo variance is prohibitive: a single S21 finds a block once in centuries on expectation. Pool only, preferably FPPS. The math is in solo vs pool for a small farm.
Solo via solo pools (solo.ckpool and similar) is a fair-odds lottery: 0.5-2% infrastructure fee, the whole reward is yours if you hit. Routing 5-10% of hashrate there as a lottery ticket is defensible; your main capacity is not.
Hardware before pool: ASIC efficiency (W/TH) dominates the economics; the pool comes second. Compare models in the miner ranking.
The algorithm: 6 steps
- Decide your variance tolerance: need smooth payouts - FPPS; can wait - PPLNS.
- Compute net BTC/day for your hardware and power price in the calculator.
- Shortlist 3-4 pools with the right scheme and a threshold your hashrate clears in days.
- Ping each pool’s stratum servers.
- Split hashrate across two pools for two weeks; compare actual payout per TH/s.
- Keep the winner; leave the runner-up configured as failover.
Common mistakes
- Choosing on the advertised fee without checking the payout base.
- PPLNS on a small pool with small hashrate: variance squared.
- Ignoring the payout threshold: six months of mining stuck in limbo.
- One pool, no failover: a day of downtime on every outage.
- Comparing pools across different weeks: difficulty and mempool shift the base - compare in parallel.
FAQ
What is the best pool for a beginner?
A large FPPS pool with a low payout threshold and servers near you. The POOL BTC ranking on the home page shows concrete options for your hashrate and power price.
FPPS or PPLNS - which earns more?
Over a long horizon PPLNS usually nets 1-2% more thanks to lower fees, but demands tolerance for weeks without payouts. FPPS buys predictability for an extra fee.
Can I mine on two pools at once?
Yes - most ASICs support a primary pool plus failover. To compare income, split capacity 50/50 for two weeks and measure payout per TH/s.
How often should I switch pools?
Not without a reason. Revisit after material changes: a fee increase, a payout scheme change, or actual payouts consistently lagging hashprice.
What is pool luck and does it matter?
Luck is found blocks versus expected blocks. Over short windows it is noise; systematic luck below 95% over a long window is a reason to look closer.
Does the pool pay out network transaction fees?
Depends on the scheme: FPPS includes network fees in the payout, classic PPS does not. That gap is exactly the difference between the advertised and the effective fee.